Vāgdhenu

Vāgdhenu

A Sanskrit chant text-to-speech demo that takes a verse in any of several Indic scripts, lets you pick the meter, and renders it as audio. Runs on Hugging Face ZeroGPU.

Made by prathoshap on Hugging Face. These notes are Mindket's, written from its Space page and README; the app's own files and license are on Hugging Face.

About

What it is: Vāgdhenu is a text-to-speech demo for Sanskrit chant. You enter a verse in one of nine Indic scripts, including Devanagari, Kannada and Grantha, and the app detects the script for you. You pick the meter, called chandas, and it renders the verse as chanted audio. It runs on Hugging Face ZeroGPU.

Who it suits: It suits people who study, recite or build tools for Sanskrit and Purāṇic verse and want to hear a verse in its meter.

Title: Who might use this demo
Industry | Who uses it | How they use it | Example
Education | Sanskrit students | Hear a verse chanted in its meter | Enter a verse and pick its chandas
Cultural heritage | Archivists | Voice written verses from different scripts | Render a verse written in Grantha
Speech technology | Developers | Try metered chant synthesis | Compare the same verse in two meters

If you want spoken text in everyday Hindi or another modern language, you can skip it, since it is built for metered chant.

Where to use it: You can use it online or run a copy yourself.

• Web browser: the Space's page on Hugging Face, with nothing to install
• Windows, macOS or Linux: a local copy that runs with Python
• Hugging Face Spaces: where the hosted demo runs, on ZeroGPU hardware

Title: What the demo is good for
Use case | Who | What they get back
Hear a verse chanted | Students | Audio of the verse in the meter you chose
Voice mixed-script texts | Archivists | Audio from verses written in different Indic scripts
Try chant synthesis | Developers | A look at how a metered chant model sounds

How to set it up: The hosted demo needs nothing. A local copy needs Python and pip.

• Python, with pip
• The Python packages that requirements.txt lists

1. Install the huggingface_hub package.
2. Download the Space's files into a folder.
3. Move into that folder.
4. Install Gradio 5.49.1.
5. Install the packages from requirements.txt.
6. Start the app with Python.

pip install -U huggingface_hub
hf download prathoshap/vagdhenu-demo --repo-type space --local-dir vagdhenu-demo
cd vagdhenu-demo
pip install gradio==5.49.1
pip install -r requirements.txt
python app.py

How to use it: Open the Space online, or start your copy with the command below. Then enter a verse, choose the meter and render it.

python app.py

Title: The steps on the page
Tool | What it does
Verse input | Takes a verse in any supported Indic script and detects the script itself
Meter choice | Lets you pick the chandas the verse should be chanted in
Render | Produces the chanted audio

Example: You paste a verse in Devanagari, choose its meter and render. The page returns audio of the verse chanted in that meter. The first request is slower, and later ones are quicker.

For newcomers and experts: If you are new to this, start with a short verse whose meter you know. Expect the first render to take longer than the rest.

If you know the field, the weights are pulled at runtime from the project's own repository. The first request downloads them and loads a DiT and BigVGAN onto the GPU, and later requests reuse the loaded models. The Space also names IndicF5.

You enter a verse in an Indic script
The app detects the script
You pick the meter
The models render the chant

Good to know: The first request is slow because it loads the models.

• First request: it downloads weights and loads the models, and later requests reuse them
• Scripts: nine Indic scripts are detected automatically
• Hosting: the hosted copy runs on ZeroGPU hardware
• License: apache-2
• Cost to run: none named
• Data it reaches: the verse you enter, plus the model weights it downloads from the project's repository
• Not for: speech in modern languages, since it renders metered Sanskrit chant

On Hugging Face, 13 November 2026

• Likes: 11
• Built with: gradio
• License: apache-2.0

Links

• Space: https://huggingface.co/spaces/prathoshap/vagdhenu-demo
• Files: https://huggingface.co/spaces/prathoshap/vagdhenu-demo/tree/main

Try it

Open the Space in your browser, or copy it to your own Hugging Face account from its page. Nothing in this kit runs by itself.